Training for Productivity: An Experimental Evaluation of Civil Service Reform in Ghana

Last registered on May 06, 2019


Trial Information

General Information

Training for Productivity: An Experimental Evaluation of Civil Service Reform in Ghana
Initial registration date
September 24, 2017

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
September 25, 2017, 3:15 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
May 06, 2019, 11:58 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.



Primary Investigator

University College London (UCL)

Other Primary Investigator(s)

PI Affiliation
Oxford University
PI Affiliation
World Bank

Additional Trial Information

In development
Start date
End date
Secondary IDs
The Training for Productivity (TFP) project aims to improve the Ghanaian Civil Service’s capacity by designing, implementing and evaluating novel individual and group-based training methods for bureaucrats. The training programs will be experimentally implemented as part of a standard package of training sessions coordinated by the Office of the Head of Civil Service (OHCS) and the Civil Service Training Centre (CSTC) of Ghana. The project findings will feed directly into the work of the OHCS, which requested the project and with whom we are working closely. As the trainings are being delivered by Civil Service personnel and through existing training systems, the research findings will have external validity and it will be straightforward to scale-up and sustain them.
External Link(s)

Registration Citation

Rasul, Imran, Daniel Rogger and Martin Williams. 2019. "Training for Productivity: An Experimental Evaluation of Civil Service Reform in Ghana." AEA RCT Registry. May 06.
Former Citation
Rasul, Imran, Daniel Rogger and Martin Williams. 2019. "Training for Productivity: An Experimental Evaluation of Civil Service Reform in Ghana." AEA RCT Registry. May 06.
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details


Our evaluation will have two intervention arms, each of which will affect the performance of the civil service. The first treatment arm will target individual civil servants through attendance of a TFP workshop; the second treatment arm will include every member of selected divisions (these are departments of a civil service organization) who will attend group training sessions. The design will cross randomize these trainings. The control groups will be the individuals and organizations that are not treated by either intervention.

The two training interventions are briefly summarized as follows:

o Training for improving productivity at the individual level: Exercises that nurture participant’s ability to conceptualize their path to implementing best practices in their divisions. These include working through diagnostic tools that identify problems within the division and outline underlying constraints that are a barrier to resolution. The activities under the individual training then include the creation of workplans for resolving these barriers and role-playing of different members of the team to assess their potential responses.

o Training for improving productivity at the organizational team level: Division-based training that aims to generate a set of common norms and expectations over reforms within the division. Exercises include creating a charter of agreed reforms for the division, undertaking problem solving exercises, role-playing and simulations of the process of implementing those reforms within the division.
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Working closely with the Office of the Head of the Civil Service, we have access to the universe of administrative data collected on public officials by the centralized agencies of government. We will complement this data collection with surveys and potentially lab experiment sessions.
Primary Outcomes (explanation)
Outcomes will be collected from a combination of survey questions. e.g. Perry Public Service Motivation, and administrative data (e.g. initiation and completion rates for divisional projects).

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
To implement an intervention cost effectively on a large scale, we take advantage of the Ghanaian Civil Service’s existing “Scheme of Service” (SoS) training system, under which all senior bureaucrats cycle through a two week training program at the government-run CSTC once every three to five years, depending on grade. As training completion is necessary to qualify for a promotion interview, the take-up rate for training is almost 100%. Approximately 700 bureaucrats will cycle through the SoS trainings at the CSTC in 2017, comprising one-quarter of the entire Service.

The SoS training is a standardized curriculum (available on request with associated materials) that outlines the standards and best practices of a professional civil servant. It is based on international best practices for public sector officials. It has been criticized for being too abstract and theoretical, with no lessons on how to apply the taught principles to a Ghanaian civil servants actual work. It is this concern that our intervention seeks to mitigate. Importantly, the Ghanaian SoS trainings are similar to many of those in the African region (World Bank, 2016). Our intervention is therefore designed to be a complementary addition to such trainings across the Africa region. By trying to identifying an intervention that could significantly amplify the productivity impacts of existing service trainings, we believe our experiment has substantial policy significance.

Our first experimental training treatment arm (T1) introduces a new training module on applied productivity techniques and teamwork into the individual-based SoS trainings. This new module comprises three two-hour sessions: 1) Understanding Productivity in Teams; 2) Management and Productivity in Ghana’s Civil Service; and 3) Applied Techniques for Productivity (all based on Oxford Policy Managements standard curriculum for public sector organizational change). This module replaces CSTC’s existing sessions on productivity and teamwork, which have been particularly highlighted by senior officials in Ghana for being too abstract and theoretical. The new module will be randomly included in half of all SoS individual trainings throughout the year. Individual officers will be allocated to the existing (control) or treatment productivity on arrival at the training centre.

Our second training treatment arm (T2) is a one-day follow-up training with the entire division of selected SoS participants. This group-based training will be held within three weeks after the individual bureaucrat has concluded their SoS training. The objective of these group training sessions is to increase the likelihood of new ideas from the individual-based training actually being implemented and learned from in organizations: since the focus of the trainings is on team productivity, uptake of practices and change in divisional culture may be higher when all members of the division undergo the complementary group-based training together. These trainings will be conducted at the Office of the Head of Civil Service rather than CSTC, but the themes of the trainings will be identical to T1, simply focused on group dynamics.

Approximately 100 SoS participants will be randomly selected from both T1 and non-T1 sessions, and letters sent (by the Head of Civil Service and Principal of CSTC) requiring the attendance of their entire division at the follow-up training on a mutually agreed date.
Experimental Design Details
Randomization Method
Randomization done in office by a computer.
Randomization Unit
Individuals and divisions will be the units assigned to the two dimensions of treatment.
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
There are around 20 individual level training sessions through the year. Divisions are treated independently of each other.
Sample size: planned number of observations
Our outcome measures and therefore units of analysis are: 1) individuals; and 2) divisions. The CSTC provides SoS training to all members of the civil service (except Directors) once every 3-5 years. While we are still working with CSTC and OHCS to construct a precise estimate of the number of individuals that will be eligible to attend SoS trainings in 2017, we know that in 2015 700 individuals attended. We therefore conservatively estimate that 25% of non-directors, or 673 individuals from 173 divisions, will attend the course in 2017. Divisions are the largest sub-organizational grouping in the Civil Service, and are the primary team within which bureaucrats work. There are 246 unique organization-divisions in Ghana’s Civil Service. The mean division has 11 members. While most divisions are small there is a tail of 30 divisions with more than 20 members - for purposes of treatment, we sub-divide these large divisions into units for <20 members, giving a total of 358 effective divisions for the purpose of our treatment size calculations.
Sample size (or number of clusters) by treatment arms
We expect around 700 individuals to be subject to the T1 intervention (with half being assigned to treatment, half to control). We expect the number of divisions treated to cover around 40% of all civil servants.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
BSG, Oxford University
IRB Approval Date
IRB Approval Number
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information


Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information


Is the intervention completed?
Data Collection Complete
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials